sites available for hydrogen gas bubble nucleation. Other techniques 

 for increasing the cathode surface area, such as scratching or sanding, 

 produce similar, but less dramatic, effects. 



Electrolyte salinity has a pronounced effect on the reaction rate. 

 Figure 12 shows seawater (34 o/oo salinity) as a standard; however, salt- 

 saturated seawater will produce a much higher reaction rate. This high 

 rate will continue for only a short time (about 1 hour), because the 

 reaction product, a jelly-like substance, will increase the electrolyte 

 density and thereby reduce the reaction rate to an unusable level. 

 Very few ions are present in an extremely low salinity electrolyte; 

 therefore, the reaction rate of the cell will be negligible in terms of 

 heat production. In general, the salinity and buffering quality of 

 seawater combine to create the highest long-term reaction rate possible. 



The thickness and surface condition of the cathode can easily be 

 controlled to provide the desired power output. In general, the cell is 

 subject to operation within the normal local salinity range. However, 

 where required, rapid heat-up can be achieved by ''spiking'* the initial 

 charge of seawater with salt. Normal electrolyte /seawater exchange will 

 flush the jelly-like products from the cell chamber, reduce the salinity, 

 and return the cell to the normal power level. 



Reaction Control. The development of a reliable and simple control 

 system for maintaining a constant power output was explored to reduce 

 the size and weight of the cell to less than that of the fixed-plate 

 cell. Table 4 summarizes the results of these investigations. The most 

 promising method appears to be of the inert cone spacer (Figure 13), 

 which uses the reaction itself ti control spacing. As of now, a simpli- 

 fied method for assembling these cells and also providing the short 

 circuit has not been developed. There are two major problems in each of 

 the control techniques. First, cell construction becomes too compli- 

 cated, and, second, the weight and volume of a cell plus control system 

 exceed those of a cell designed to provide the same delivered final 

 power without control. 



A variety of other tests were performed in attempts to control or 

 modify the reaction rate. These tests are summarized in Table 4. 

 None of the test results showed improvements significant enough to cause 

 modification of the basic dual-plate cell configuration. 



Bi-Polar Electrode 



One of the disadvantages of the dual-plate cell is holes start to 

 form in the anode during the last hour of the reaction. The holes 

 decrease fhe current-carrying cross section of the anode and reduce the 

 active anode area. The result is reduced cell power. To minimize these 

 effects in the dual-plate cell the anodes must be slightly thicker than 

 actually needed. This insures enough active surface area will remain to 

 provide the required power for the desired duration. 



. . , i ,,.,, ■■■■'-f-^T . r'f nii.r.rtt. f-' n-'ii V i.' V i.i - , - ■ , i 



.-,,,. ., , ,.. i i,...vm5!^..i.,.a.,.a^..; , .-.a 



